Uveal Melanoma in Young Patients

Objective  To study the clinical profile of young patients with uveal melanoma.

Design  Retrospective case-control series.

Setting  Tertiary referral center.

Patients  Data on 63 patients aged 20 years or younger with uveal melanoma were reviewed for clinical profile and association with oculo(dermal) melanocytosis, familial uveal melanoma, dysplastic nevus syndrome, cutaneous melanoma, and other second malignant neoplasms.

Results  Of 8000 patients with uveal melanoma, 63 (0.8%) were found in patients who were 20 years of age or younger. The median age at diagnosis was 16 years, and the youngest patient was 3 years old. Sixty-two patients (98%) were white, and uveal melanoma was unilateral in all cases. Seven patients (11%) had oculo(dermal) melanocytosis. Two patients (3%) had dysplastic nevi syndrome, and personal history of cutaneous melanoma was observed in 1 patient (2%). No other second cancers were present in any patient. The 5- and 15-year posttreatment survival estimates were 0.95 (95% confidence interval, 0.87-1.00) and 0.77 (95% confidence interval, 0.52-1.00), respectively.

Conclusions  Uveal melanoma is rare in children or teenagers. It occurs in a heterogeneous group displaying various associations, especially with oculo(dermal) melanocytosis. Oculo(dermal) melanocytosis is 9 times (95% confidence interval, 3.6-22.8) more common in young patients with uveal melanoma than in the general population with uveal melanoma. Young patients with uveal melanoma have short-term (5-year) survival better than that of adults, but the long-term (15-year) survival is similar to that of adults.

UVEAL MELANOMA rarely occurs in young people. There are few reports describing clinical and histopathological features of uveal melanoma in young patients.^1-8 Most young patients with uveal melanoma are pubertal, although uveal melanoma can even be present at birth.^9,10 The clinical features and management of uveal melanoma in young patients are similar to those of uveal melanoma in adults.^7,8,11

Environmental, host, and genetic factors involved in the pathogenesis of uveal melanoma have not been completely elucidated.¹² In rare instances, uveal melanoma occurs in the presence of oculo (dermal) melanocytosis, neurofibromatosis type 1, and dysplastic nevus syndrome (familial atypical mole and melanoma syndrome), suggesting that, at least in some cases, there may be an inherited predisposition to develop uveal melanoma.¹³ The concept of genetic predisposition to uveal melanoma is further supported by the occurrence of uveal melanoma in some families (familial uveal melanoma).^14-17

In general, patients with a cancer predisposition syndrome tend to develop cancer at an earlier age.¹⁸ Observations have been previously reported on our patients with uveal melanoma who demonstrated features of cancer predisposition such as familial uveal melanoma,¹⁴ bilateral primary uveal melanoma,¹⁹ and uveal melanoma with phenotypic associations such as oculo(dermal) melanocytosis²⁰ and familial atypical mole and melanoma syndrome syndrome.²¹

In the present report, we detail the clinical profile and analysis of 63 young patients with uveal melanoma. We specifically investigated the presence of any clinically detectable genetic predisposition in these patients and the relationship of uveal melanoma to familial uveal melanoma, oculo(dermal) melanocytosis, dysplastic nevus syndrome, cutaneous melanoma, and other systemic malignant neoplasms.

The computerized database of the Oncology Service at the Wills Eye Hospital, Philadelphia, Pa, was reviewed, and patients with primary uveal melanoma diagnosed at the age of 20 years or younger were included in the study. All patients were diagnosed as having uveal melanoma based on ophthalmoscopic, fluorescein angiographic, and ultrasonographic appearance and confirmed histopathologically in eyes treated by enucleation or resection. The clinical data, including age at diagnosis, tumor location, associated ocular findings, method of treatment, and histopathological features, were collected from review of the medical chart. Information regarding presence of any systemic disease, metastatic disease, or personal or family history of malignant neoplasm in first-degree relatives was updated at the last follow-up visit and by a questionnaire mailed to the last known address. Specifically, data regarding predisposing conditions such as familial uveal melanoma, oculo(dermal) melanocytosis, familial atypical mole and melanoma syndrome, cutaneous melanoma, and other systemic malignant neoplasm were recorded. Prevalence of oculo(dermal) melanocytosis in our series was compared with previously published prevalence rates of oculo(dermal) melanocytosis in the general uveal melanoma population by means of Fisher exact test (2-tailed).²² The Kaplan-Meier product-limit method was used to compute the survival estimates for the sample, as it did not require a priori assumptions about the distribution of the survival function.²³

Of the 8000 patients with uveal melanoma treated on the Oncology Service at Wills Eye Hospital between January 1, 1974, and August 31, 1998, we identified 63 patients (0.8%) with uveal melanoma who were diagnosed at the age of 20 years or younger (Table 1). The median age at diagnosis was 16 years (range, 3-20 years) (Figure 1). The median follow-up was 51 months (range, 1-307 months). Follow-up duration was less than 5 years in 30 patients, between 5 and 10 years in 16 patients, between 10 and 15 years in 11 patients, and more than 15 years in 6 patients. On the last follow-up, 4 patients were dead of metastatic melanoma (proved by biopsy). The 5-, 10-, and 15-year survival estimates were 0.95 (95% confidence interval [CI], 0.87-1.00), 0.89 (95% CI, 0.79-1.00), and 0.77 (95% CI, 0.52-1.00), respectively (Figure 2).

Sixty-two patients (98%) were white, 1 (2%) was Asian, and none was African American. There were 38 female patients (60%) and 25 male patients (40%). All cases were unilateral, with 34 (54%) affecting the left eye and 29 (46%) affecting the right eye. The tumors involved primarily the iris in 16 cases (25%), ciliary body in 13 (21%), and choroid in 34 (54%). The treatments included enucleation in 39 patients (62%), local resection in 12 (19%), and radioactive plaque in 9 (14%). In all patients for whom tumor samples were available (51 of 63 or 81%), including 15 cases of iris melanoma, histopathological examination confirmed the diagnosis of uveal melanoma. The tumor cell type included spindle cell in 29 cases (57%), mixed cell in 17 (33%), and epithelioid cell in 5 (10%).

The only associated ocular finding was oculo(dermal) melanocytosis in 7 patients (11%). The oculo(dermal) melanocytosis was diffuse, involving all quadrants of the uvea, in 6 cases and sectoral in 1 case, with choroidal melanoma occurring in the sector with melanocytosis (Figure 3). Compared with the general population of patients with uveal melanoma, oculo(dermal) melanocytosis was 9 times (95% CI, 3.6-22.8) more common in young patients with uveal melanoma (P<.001).²² The associated systemic findings included familial uveal melanoma (1 case), dysplastic nevus syndrome (2 cases), and history of cutaneous melanoma (1 case). None of the patients had a personal history of prior or subsequent development of a second primary malignant neoplasm.

The patient with familial uveal melanoma had a history of histopathologically confirmed uveal melanoma in the maternal grandmother (Figure 4). She also had multiple cutaneous biopsy specimens that were confirmed by histopathological examination to be dysplastic, consistent with the diagnosis of dysplastic nevus syndrome.²⁴ The patient with a personal history of cutaneous melanoma is described in detail below.

A 16-year-old white girl noted visual disturbance of a few weeks' duration and was referred to the Oncology Service at Wills Eye Hospital for management of a suspected choroidal melanoma of the left eye in February 1998. Six months before this examination, she had been diagnosed as having a choroidal nevus.

At age 6 years, a "mole" had been noted on the top of the patient's great toe. Initial biopsy showed it to be a nevus. Subsequent local recurrence led to further excision and cryoablation. Finally, the great toe was amputated, and the diagnosis of stage IV malignant melanoma was established. Recent systemic evaluation showed no dysplastic nevi or systemic metastasis. There was no family history of cutaneous or uveal melanoma.

On examination of the eye, visual acuity was 20/20 OD and 20/400 OS. Examination of the right eye showed normal findings. There was no heterochromia or oculo(dermal) melanocytosis in either eye. Ophthalmoscopically, a dome-shaped choroidal melanocytic lesion superior to the optic disc was noted. The lesion measured 13 × 13 mm in basal dimension and 8 mm in thickness and was associated with subtotal exudative retinal detachment (Figure 5). Ultrasonographic findings of low internal reflectivity and fluorescein angiographic findings were consistent with the diagnosis of primary choroidal melanoma. After the parents were counseled regarding various therapeutic options, enucleation was performed. Histopathological examination confirmed the lesion to be a primary uveal melanoma with predominant spindle cells.

Peripheral-blood lymphocyte DNA was extracted by means of a blood DNA extraction kit (Qiagen, Chatsworth, Calif). Exons of the CDKN2A tumor suppressor gene (also known as p16, p16^INK4a, and MTS1) were amplified by polymerase chain reaction with the use of 5 to 10 ng of germline DNA as previously described.²⁵ The polymerase chain reaction products were sequenced with an automated sequencer (model 373A; Applied Biosystems, Foster City, Calif). No abnormality was detected.

We identified 63 patients (0.8% of a total of 8000) with uveal melanoma that was diagnosed at the age of 20 years or younger, a criterion used consistently in the literature to classify young patients with uveal melanoma.^7,8 In previously published series, including the series from the Armed Forces Institute of Pathology, Washington DC, uveal melanoma in young patients has composed 0.6% to 1.6% of all cases of uveal melanoma.^7,8 The median age at diagnosis was 16 years (range, 3-20 years). More than 90% of patients were 10 years old or older, and more than 50% of patients were at least 16 years old.^7,8 None of the cases was congenital.^1-8 Our data compare favorably with previously published reports suggesting that onset of uveal melanoma in children may be influenced by onset of hormonal changes of puberty.^7,8 Many of the clinical features, such as unilaterality with equal distribution between the 2 eyes, predominant white racial distribution (62 of 63 or 98%), choroidal location (34 of 63 or 54%), and methods of treatment applied, were similar to those of adult patients with uveal melanoma.¹¹

The median follow-up was 51 months (range, 1-307 months). At last follow-up, 4 patients had died of metastatic disease. The 5-, 10-, and 15-year survival estimates for the sample were 0.95 (95% CI, 0.87-1.00), 0.89 (95% CI, 0.79-1.00), and 0.77 (95% CI, 0.52-1.00), respectively (Figure 2). In a previous report,⁸ short-term (5-year) survival of 96% was reported, similar to the present estimate of 95%. Now, with the longer follow-up in the present series, the long-term survival is estimated to be 77%, similar to the survival estimates reported by Barr and associates.⁷ Based on the small number of selected cases in our series, it appears that young patients with uveal melanoma have long-term survival (15 years) similar to that of adults and short-term survival (5 years) possibly better than that of adults.²⁶ Mortality in our series resulted exclusively from metastatic uveal melanoma, confirmed by biopsy in each case. Patients in our series demonstrated a lower mortality rate at 5 and 10 years compared with uveal melanoma–specific mortality in adults.²⁷ However, survival rates in young patients with uveal melanoma cannot be directly compared with survival data in adults with uveal melanoma, as there may be differences in baseline tumor measures between the 2 groups. Ascertainment bias caused by the referral nature of our practice may also influence our results.

Since early onset of malignant neoplasm can be a feature of genetic predisposition, we undertook the present study to investigate the role of any possible predisposing conditions in young patients with uveal melanoma. The associated ocular findings identified in young patients with uveal melanoma included oculo(dermal) melanocytosis in 7 patients (11%).^28,29 Verdaguer,⁴ in his report on young patients with uveal melanoma, observed oculo(dermal) melanocytosis in 4 of the 7 patients. Oculo(dermal) melanocytosis is a congenital, nonhereditary condition characterized by hyperpigmentation of the episclera, uvea, orbit, meninges, and periocular skin (nevus of Ota).²² In the white population, there is a strong association between oculo(dermal) melanocytosis and uveal melanoma, with the lifetime risk of developing uveal melanoma estimated to be 1 in 400.²⁰ Approximately 1.4% of adult patients with uveal melanoma have associated oculo(dermal) melanocytosis.²² Our data indicate that the prevalence of oculo(dermal) melanocytosis in young patients with uveal melanoma is 9 times (95% CI, 3.6-22.8) greater than in the general population with uveal melanoma. It has been postulated that an excess of melanocytes in the uveal tract in oculo(dermal) melanocytosis may be the basis for susceptibility to the development of uveal melanoma.²⁰ However, the age at diagnosis of uveal melanoma in patients with oculo(dermal) melanocytosis is similar to that of random uveal melanoma, with only 0.5% of patients developing uveal melanoma at age 10 years or younger.²⁰

The associated systemic findings identified in our series of uveal melanoma in young patients included familial uveal melanoma in 1, dysplastic nevus syndrome in 2, and a history of cutaneous melanoma in 1 patient. One patient had familial uveal melanoma with a history of histopathologically confirmed uveal melanoma in the maternal grandmother (Figure 3). She also underwent multiple cutaneous biopsies for suspicious cutaneous moles, and some were confirmed by histopathological examination to be dysplastic, consistent with the diagnosis of dysplastic nevus syndrome. There was no personal or family history of cutaneous melanoma. It has been reported that the median age at diagnosis of uveal melanoma in familial uveal melanoma is comparable to that in sporadic cases.^14-16 Although offspring may have an earlier onset of uveal melanoma,^14-16 none of the published cases of familial uveal melanoma have included uveal melanoma in young patients.^14-16

Included in our present series is another case of a 10-year-old child with iris melanoma and dysplastic nevi syndrome published in detail elsewhere.³⁰ Dysplastic nevus syndrome (familial atypical mole and melanoma syndrome) is an autosomal dominant condition with marked predisposition for multifocal and familial cutaneous melanoma.²⁴ Despite several reported cases of uveal melanoma in adults and children with dysplastic nevus syndrome,^31,32 the association between uveal melanoma and dysplastic nevus remains uncertain partly because of controversy regarding the criteria used for diagnosis of dysplastic nevi syndrome.^33,34 Except for 1 patient (case report) with a history of cutaneous melanoma (in the absence of dysplastic nevi), none of the patients had a previous or subsequent development of a second primary malignant neoplasm. None of the patients reported systemic illness or use of medications causing immunosuppression, a causative factor for cutaneous melanoma in children.³⁵ Perhaps longer follow-up is needed to completely exclude genetic susceptibility to develop other cancers. Some authors believe that the coexistence of primary uveal melanoma and primary cutaneous melanoma may be coincidental, whereas others believe that there is an association between uveal melanoma and cutaneous melanoma because of similar embryological and morphological features of cutaneous and uveal melanocytes.³⁶

On evaluating features of a cancer predisposition syndrome, such as development of cancer at an earlier age, bilateral involvement of paired organs, multiple primary tumors, familial occurrence of cancer, and phenotypic associations in relation to uveal melanoma, each subtype of uveal melanoma has been identified.^13,18 These rare subtypes include pediatric, bilateral, multifocal, and familial uveal melanomas. Syndromic association of uveal melanoma include oculo(dermal) melanocytosis and possible variants such as dysplastic nevus syndrome and neurofibromatosis type 1.^13,37-39 We are not aware of any reports wherein a patient with uveal melanoma had multiple features suggestive of an inherited predisposition, such as early onset, family history of uveal melanoma, and bilateral involvement. These findings suggest that inheritable genetic traits play a minor role in pathogenesis of uveal melanoma. However, patients with "atypical uveal melanoma" provide unique opportunities for molecular genetic testing, as carried out in one of our patients.

Approximately 10% of patients with cutaneous melanoma have a family history of this condition.⁴⁰ Presence of dysplastic nevi is an important risk factor for development of cutaneous melanoma in both sporadic and familial cases.⁴⁰ The cutaneous melanoma in patients with dysplastic nevi tends to occur at an earlier age, with 10% of such patients manifesting their first cutaneous melanoma before the age of 20 years.³⁵ Recent molecular genetic studies have demonstrated germline mutations of the CDKN2 A gene (chromosome 9 p21 region) in 20% of families with predisposition to cutaneous melanoma. ^40-42 The CDKN2 A gene, also known as p16^INK4a or multiple tumor suppressor gene 1 (MTS1), negatively regulates the cell cycle.^43,44 Screening of exons for mutations by direct sequencing did not disclose any abnormality in our patient (see case report). Previous investigations^8,45 have shown a similar lack of CDKN2 A gene germline mutations in 11 patients with familial and 2 with bilateral uveal melanoma, excluding CDKN2 A gene as a uveal melanoma predisposition gene in the majority of cases.

In summary, uveal melanoma rarely occurs in the young population (less than 1%). Oculo(dermal) melanocytosis is 9 times more commonly associated with uveal melanoma in young patients than in the general uveal melanoma population. Detailed history and a thorough systemic examination are recommended to detect any other predisposing condition. Long-term clinical studies and molecular genetic studies are needed to completely evaluate the possibility of a genetic predisposition in young patients with uveal melanoma.

Accepted for publication February 2, 2000.

This investigation was supported by the Eye Tumor Research Foundation, Philadelphia, Pa; Macula Foundation, New York, NY (Dr C. Shields); and the Paul Kayser Award of Merit in Retinal Research, Houston, Tex (Dr J. Shields).

Biostatistical consultation was provided by Jacqueline Cater, PhD.

Corresponding author: Arun D. Singh, MD, Oncology Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107 (e-mail: arunsingh@eyetumors.com).